Field of the Invention
The present invention relates to a three-dimensional image processing apparatus, a three-dimensional image processing method, and a three-dimensional image processing program for supporting insertion of an endoscope into a pipe line in the body such as a bronchial tube using a three-dimensional image obtained by imaging a test object.
Related Art
Recently, along with advances of an imaging apparatus, resolution of image data obtained by imaging using an imaging apparatus has been enhanced, and detailed analysis of a test object has become possible on the basis of the image data. For example, multi detector-row computed tomography (CT) is able to capture a plurality of tomographic images all at once, and is able to capture tomographic images with a thin slice thickness. As the slice thickness becomes thin, resolution of a three-dimensional image in which a plurality of tomographic images is layered in a body axis direction is enhanced, to thereby make it possible to obtain a detailed three-dimensional image. By displaying and analyzing such a three-dimensional image, it is possible to find out lesion or the like that has not easily been found so far.
As one display method using the above-described three-dimensional image, a virtual endoscope display method is used (for example, see WO2014/141968A and JP1999-120327A (JP-H11-120327A)). The virtual endoscope display method refers to a method for setting a viewpoint position inside a lumen, generating a perspective projection image on the basis of the viewpoint position, and displaying the perspective projection image. In the virtual endoscope display method, as a user sequentially changes viewpoint positions, it is possible to provide an image as if a camera of the endoscope performs imaging while moving inside the body.
Particularly, in bronchoscopy that is an inspection using an endoscope, since divergence of bronchial tubes is extremely complicated, insertion of the endoscope is performed while referring to the above-described virtual endoscope display as a “map”. In this case, since it is laborious to sequentially change the virtual endoscope display method manually to match movement of an actual endoscope, a method for estimating which place in the body a tip of the endoscope is located at to create a virtual endoscope image.
For example, WO2014/141968A proposes a method for performing a registration process of a virtual endoscope image and an endoscope image that is actually captured using an endoscope so as to estimate a tip position of the endoscope.
Here, in order to estimate the tip position of the endoscope for navigation, it is necessary to generate virtual endoscope images from a plurality of viewpoints, and to perform a registration process of the virtual endoscope images and actually captured endoscope images to select most similar images.
However, in order to enhance estimation accuracy or stability of a tip position of an endoscope, it is necessary to generate a large amount of virtual endoscope images, but a volume rendering process that is a process of generating virtual endoscope images takes a lot of time.
JP2013-192741A discloses a technique for generating a graphical structure from a three-dimensional image of bronchial tubes, and generating, with respect to each of a plurality of divergent points included in the graphical structure, a projection image obtained by projecting divergence destinations from the divergent point on a two-dimensional plane including the divergent point, but since the projection image is a schematic image indicating the divergence destinations and is not an image that correctly represents shapes or the like of holes of bronchial tubes, the projection image is not suitable for the above-described registration process with respect to the endoscope image.
An object of the present invention is to provide a three-dimensional image processing apparatus, a three-dimensional image processing method, and a three-dimensional image processing program capable of generating, through a high-speed process, an image in which a direct feature is remained so that a registration process with respect to an endoscope image can be performed.